Research On Case Matching Of Urban Bus Fire Accident

In recent years,with the expansion of the city circle,there are more and more buses in major cities across the country,the bus lines become more complex and the coverage are more widely,at the same time the bus fire accidents occur more frequently.In many bus fire accidents,heavy casualties always happened,each accident not only causes the destruction of vehicles,but also hinders the normal traffic order and brings adverse effects to the people.In the event of a bus fire,the evacuation of passengers in the bus is much more difficult than the average place.The smoke and fire generated by the accident and the crowded trampling caused by the panic of the passengers,.et al will bring more casualties on the basis of difficult evacuation.As the main carrier of urban public transport,the safety of the bus must be paid more attention.How to excavate the accident information from the historical accident and sum up the lessons learned,so as to formulate the bus fire accident disposal strategy and prevention measures,are significant to ensure the safety of the public.In order to better summarize the lessons of past accidents,this thesis conducted case matching research on bus accident based on CBR.On the basis of using frame representation model to express bus fire accident case,preliminary hazard analysis(PHA)method was used to analyze the cause and consequence of the accident,and the accident characteristic index system was developed according to the national accident statistics standard,then this research collected 110 bus fire accidents between 2005 and 2014 through the Internet,news reports and so on,after organizing accident information,finally the bus fire accident case library was built and the case database was got by gathering statistics.After comparing the merits and demerits of the case retrieval model,this thesis designed the case retrieval model of the bus accident by combining the information entropy and the K-NN proximity algorithm.Through the model calculation,the similar history cases of the target accident case "1·5 Yinchuan bus arson case were found.The accuracy of the retrieval model is verified by comparing the similarity of basic information between similar cases and the target case.After comparing the consequences of these accident cases and analyzing the causes of these differences,the man-made bus fire accident disposal strategies and prevention measures were developed in this thesis.After further analyzing the data of the accident case database,the general law of the accident is obtained,and finally,more targeted prevention and control policies were developed.